Pull x86 mm updates from Ingo Molnar:
"Various x86 low level modifications:
- preparatory work to support virtually mapped kernel stacks (Andy
Lutomirski)
- support for 64-bit __get_user() on 32-bit kernels (Benjamin
LaHaise)
- (involved) workaround for Knights Landing CPU erratum (Dave Hansen)
- MPX enhancements (Dave Hansen)
- mremap() extension to allow remapping of the special VDSO vma, for
purposes of user level context save/restore (Dmitry Safonov)
- hweight and entry code cleanups (Borislav Petkov)
- bitops code generation optimizations and cleanups with modern GCC
(H. Peter Anvin)
- syscall entry code optimizations (Paolo Bonzini)"
* 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (43 commits)
x86/mm/cpa: Add missing comment in populate_pdg()
x86/mm/cpa: Fix populate_pgd(): Stop trying to deallocate failed PUDs
x86/syscalls: Add compat_sys_preadv64v2/compat_sys_pwritev64v2
x86/smp: Remove unnecessary initialization of thread_info::cpu
x86/smp: Remove stack_smp_processor_id()
x86/uaccess: Move thread_info::addr_limit to thread_struct
x86/dumpstack: Rename thread_struct::sig_on_uaccess_error to sig_on_uaccess_err
x86/uaccess: Move thread_info::uaccess_err and thread_info::sig_on_uaccess_err to thread_struct
x86/dumpstack: When OOPSing, rewind the stack before do_exit()
x86/mm/64: In vmalloc_fault(), use CR3 instead of current->active_mm
x86/dumpstack/64: Handle faults when printing the "Stack: " part of an OOPS
x86/dumpstack: Try harder to get a call trace on stack overflow
x86/mm: Remove kernel_unmap_pages_in_pgd() and efi_cleanup_page_tables()
x86/mm/cpa: In populate_pgd(), don't set the PGD entry until it's populated
x86/mm/hotplug: Don't remove PGD entries in remove_pagetable()
x86/mm: Use pte_none() to test for empty PTE
x86/mm: Disallow running with 32-bit PTEs to work around erratum
x86/mm: Ignore A/D bits in pte/pmd/pud_none()
x86/mm: Move swap offset/type up in PTE to work around erratum
x86/entry: Inline enter_from_user_mode()
...
Pull perf updates from Ingo Molnar:
"With over 300 commits it's been a busy cycle - with most of the work
concentrated on the tooling side (as it should).
The main kernel side enhancements were:
- Add per event callchain limit: Recently we introduced a sysctl to
tune the max-stack for all events for which callchains were
requested:
$ sysctl kernel.perf_event_max_stack
kernel.perf_event_max_stack = 127
Now this patch introduces a way to configure this per event, i.e.
this becomes possible:
$ perf record -e sched:*/max-stack=2/ -e block:*/max-stack=10/ -a
allowing finer tuning of how much buffer space callchains use.
This uses an u16 from the reserved space at the end, leaving
another u16 for future use.
There has been interest in even finer tuning, namely to control the
max stack for kernel and userspace callchains separately. Further
discussion is needed, we may for instance use the remaining u16 for
that and when it is present, assume that the sample_max_stack
introduced in this patch applies for the kernel, and the u16 left
is used for limiting the userspace callchain (Arnaldo Carvalho de
Melo)
- Optimize AUX event (hardware assisted side-band event) delivery
(Kan Liang)
- Rework Intel family name macro usage (this is partially x86 arch
work) (Dave Hansen)
- Refine and fix Intel LBR support (David Carrillo-Cisneros)
- Add support for Intel 'TopDown' events (Andi Kleen)
- Intel uncore PMU driver fixes and enhancements (Kan Liang)
- ... other misc changes.
Here's an incomplete list of the tooling enhancements (but there's
much more, see the shortlog and the git log for details):
- Support cross unwinding, i.e. collecting '--call-graph dwarf'
perf.data files in one machine and then doing analysis in another
machine of a different hardware architecture. This enables, for
instance, to do:
$ perf record -a --call-graph dwarf
on a x86-32 or aarch64 system and then do 'perf report' on it on a
x86_64 workstation (He Kuang)
- Allow reading from a backward ring buffer (one setup via
sys_perf_event_open() with perf_event_attr.write_backward = 1)
(Wang Nan)
- Finish merging initial SDT (Statically Defined Traces) support, see
cset comments for details about how it all works (Masami Hiramatsu)
- Support attaching eBPF programs to tracepoints (Wang Nan)
- Add demangling of symbols in programs written in the Rust language
(David Tolnay)
- Add support for tracepoints in the python binding, including an
example, that sets up and parses sched:sched_switch events,
tools/perf/python/tracepoint.py (Jiri Olsa)
- Introduce --stdio-color to set up the color output mode selection
in 'annotate' and 'report', allowing emit color escape sequences
when redirecting the output of these tools (Arnaldo Carvalho de
Melo)
- Add 'callindent' option to 'perf script -F', to indent the Intel PT
call stack, making this output more ftrace-like (Adrian Hunter,
Andi Kleen)
- Allow dumping the object files generated by llvm when processing
eBPF scriptlet events (Wang Nan)
- Add stackcollapse.py script to help generating flame graphs (Paolo
Bonzini)
- Add --ldlat option to 'perf mem' to specify load latency for loads
event (e.g. cpu/mem-loads/ ) (Jiri Olsa)
- Tooling support for Intel TopDown counters, recently added to the
kernel (Andi Kleen)"
* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (303 commits)
perf tests: Add is_printable_array test
perf tools: Make is_printable_array global
perf script python: Fix string vs byte array resolving
perf probe: Warn unmatched function filter correctly
perf cpu_map: Add more helpers
perf stat: Balance opening and reading events
tools: Copy linux/{hash,poison}.h and check for drift
perf tools: Remove include/linux/list.h from perf's MANIFEST
tools: Copy the bitops files accessed from the kernel and check for drift
Remove: kernel unistd*h files from perf's MANIFEST, not used
perf tools: Remove tools/perf/util/include/linux/const.h
perf tools: Remove tools/perf/util/include/asm/byteorder.h
perf tools: Add missing linux/compiler.h include to perf-sys.h
perf jit: Remove some no-op error handling
perf jit: Add missing curly braces
objtool: Initialize variable to silence old compiler
objtool: Add -I$(srctree)/tools/arch/$(ARCH)/include/uapi
perf record: Add --tail-synthesize option
perf session: Don't warn about out of order event if write_backward is used
perf tools: Enable overwrite settings
...
Pull RCU updates from Ingo Molnar:
"The main changes in this cycle were:
- documentation updates
- miscellaneous fixes
- minor reorganization of code
- torture-test updates"
* 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (30 commits)
rcu: Correctly handle sparse possible cpus
rcu: sysctl: Panic on RCU Stall
rcu: Fix a typo in a comment
rcu: Make call_rcu_tasks() tolerate first call with irqs disabled
rcu: Disable TASKS_RCU for usermode Linux
rcu: No ordering for rcu_assign_pointer() of NULL
rcutorture: Fix error return code in rcu_perf_init()
torture: Inflict default jitter
rcuperf: Don't treat gp_exp mis-setting as a WARN
rcutorture: Drop "-soundhw pcspkr" from x86 boot arguments
rcutorture: Don't specify the cpu type of QEMU on PPC
rcutorture: Make -soundhw a x86 specific option
rcutorture: Use vmlinux as the fallback kernel image
rcutorture/doc: Create initrd using dracut
torture: Stop onoff task if there is only one cpu
torture: Add starvation events to error summary
torture: Break online and offline functions out of torture_onoff()
torture: Forgive lengthy trace dumps and preemption
torture: Remove CONFIG_RCU_TORTURE_TEST_RUNNABLE, simplify code
torture: Simplify code, eliminate RCU_PERF_TEST_RUNNABLE
...
Here is the big Staging and IIO driver update for 4.8-rc1.
We ended up adding more code than removing, again, but it's not all that
bad. Lots of cleanups all over the staging tree, and new IIO drivers,
full details in the shortlog.
All of these have been in linux-next for a while with no reported issues.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iFYEABECABYFAleVPQQPHGdyZWdAa3JvYWguY29tAAoJEDFH1A3bLfsplRgAniG6
jfPnvlHhl70T5HsGJzrc7VS9AKCBQ5x0gzTNxo2nnGfPmR8CVEH7Bg==
=0/6X
-----END PGP SIGNATURE-----
Merge tag 'staging-4.8-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging
Pull staging and IIO driver updates from Greg KH:
"Here is the big Staging and IIO driver update for 4.8-rc1.
We ended up adding more code than removing, again, but it's not all
that bad. Lots of cleanups all over the staging tree, and new IIO
drivers, full details in the shortlog.
All of these have been in linux-next for a while with no reported
issues"
* tag 'staging-4.8-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging: (417 commits)
drivers:iio:accel:mma8452: removed unwanted return statements
drivers:iio:accel:mma8452: added cleanup provision in case of failure.
iio: Add iio.git tree to MAINTAINERS
iio:st_pressure: clean useless static channel initializers
iio:st_pressure:lps22hb: temperature support
iio:st_pressure:lps22hb: open drain support
iio:st_pressure: temperature triggered buffering
iio:st_pressure: document sampling gains
iio:st_pressure: align storagebits on power of 2
iio:st_sensors: align on storagebits boundaries
staging:iio:lis3l02dq drop separate driver
iio: accel: st_accel: Add lis3l02dq support
iio: adc: add missing of_node references to iio_dev
iio: adc: ti-ads1015: add indio_dev->dev.of_node reference
iio: potentiometer: Fix typo in Kconfig
iio: potentiometer: mcp4531: Add device tree binding
iio: potentiometer: mcp4531: Add device tree binding documentation
iio: potentiometer: mcp4531: Add support for MCP454x, MCP456x, MCP464x and MCP466x
iio:imu:mpu6050: icm20608 initial support
iio: adc: max1363: Add device tree binding
...
The curly braces are missing here so we print stuff unintentionally.
Fixes: 9da4714a2d ('slub: slabinfo update for cmpxchg handling')
Link: http://lkml.kernel.org/r/20160715211243.GE19522@mwanda
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Christoph Lameter <cl@linux.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Colin Ian King <colin.king@canonical.com>
Cc: Laura Abbott <labbott@fedoraproject.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
There are no parentheses around this macro and it causes a problem when
we do:
index = rand() % THRASH_SIZE;
Link: http://lkml.kernel.org/r/20160715210953.GC19522@mwanda
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add automated test for is_printable_array function.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Pirko <jiri@mellanox.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: http://lkml.kernel.org/r/1468685480-18951-4-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
It's used from 2 objects in perf, so it's better to keep just one copy.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Pirko <jiri@mellanox.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: http://lkml.kernel.org/r/1468685480-18951-3-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Jirka reported that python code returns all arrays as strings. This
makes impossible to get all items for byte array tracepoint field
containing 0x00 value item.
Fixing this by scanning full length of the array and returning it as
PyByteArray object in case non printable byte is found.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Reported-and-Tested-by: Jiri Pirko <jiri@mellanox.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: http://lkml.kernel.org/r/1468685480-18951-2-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Warn unmatched function filter correctly instead of warning
"symbol-loading error", since that can be a filter issue.
From the technical point of view, this adds a filter chech in map__load
and if there is a filter, it returns -2 (filter-out), instead of -1
(error), and perf-probe checks it and change message.
E.g. without this fix:
# perf probe -F rt_sp*
no symbols found in [kernel.kallsyms], maybe install a debug package?
Failed to load symbols in kernel
With this fix:
# perf probe -F rt_sp*
no symbols passed the given filter.
Failed to find symbols matched to "rt_sp*"
Error: Failed to show functions.
Reported-and-Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/146885835596.16106.2293540792775552481.stgit@devbox
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
In some cases it's necessry to figure out the map-local index of a given
Linux logical CPU ID. Add a new helper, cpu_map__idx, to acquire this.
As the logic is largely the same as the existing cpu_map__has, this is
rewritten in terms of the new helper.
At the same time, add the inverse operation, cpu_map__cpu, which yields
the logical CPU id for a map-local index. While this can be performed
manually, wrapping this in a helper can make code more legible.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1468577293-19667-3-git-send-email-mark.rutland@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
In create_perf_stat_counter, when a target CPU has not been provided, we
call __perf_evsel__open with empty_cpu_map, and open a single FD per
thread. However, in read_counter we assume that we opened events for the
product of threads and CPUs described in the evsel's cpu_map.
Thus, if an evsel has a cpu_map with more than one entry, we will
attempt to access FDs that we didn't open. This could result in a number
of problems (e.g. blocking while reading from STDIN if the fd memory
happened to be initialised to zero).
This is problematic for systems were a logical CPU PMU covers some
arbitrary subset of CPUs. The cpu_map of any evsel for that PMU will be
initialised based on the cpumask exposed through sysfs, even if the user
requests per-thread events.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Kan Liang <kan.liang@intel.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/r/1468577293-19667-2-git-send-email-mark.rutland@arm.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We were also using this directly from the kernel sources, the two last
cases, fix it.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-7o14xvacqcjc5llc7gvjjyl8@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
It hasn't been used since we made tools/ self sufficiente wrt list.h.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Fixes: d1b39d41eb ("tools: Make list.h self-sufficient")
Link: http://lkml.kernel.org/n/tip-w20ueqlf22kh7ctjqo0zjpig@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
copy some more kernel files accessed from tools/, check for drift.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-omz8xdyvvxgjiuqzwj6ecm6j@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
No need to copy it to a detached tarball as they aren't used anymore
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-lopmaqi439ke10g1j9cxrxwt@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Not used anymore, remove one more file referencing kernel sources, i.e.
outside of tools/
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-ykfjt3t8l0npxfwmekiwwyu6@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Not used anymore. This also stops include linux/swab.h directly
from the kernel sources, remove that reference from the MANIFEST.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
It uses the likely/unlikely macros, so need to include
<linux/compiler.h>.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-p0xrhgbkicsii9ohmhhprqpi@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The 'info.e_machine' struct member is an uint16_t so 'm' is never less
than zero. It looks like this was maybe left over code from earlier
versions so I've just removed it.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: kernel-janitors@vger.kernel.org
Link: http://lkml.kernel.org/r/20160715210836.GB19522@mwanda
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
It doesn't change the runtime behavior, but my static checker complains
that curly braces were intended.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: kernel-janitors@vger.kernel.org
Link: http://lkml.kernel.org/r/20160715210712.GA19522@mwanda
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
gcc version 4.1.2 20080704 (Red Hat 4.1.2-55) barfs with:
CC /tmp/build/objtool/builtin-check.o
cc1: warnings being treated as errors
builtin-check.c: In function 'cmd_check':
builtin-check.c:667: warning: 'prev_rela' may be used uninitialized in this function
mv: cannot stat `/tmp/build/objtool/.builtin-check.o.tmp': No such file or directory
make[1]: *** [/tmp/build/objtool/builtin-check.o] Error 1
Init it to NULL to silence it.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-qolo31rl2ojlwj1lj9dhemyz@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So that it can find asm/bitsperlong.h to get the __BITS_PER_LONG
definition.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-pr3pvskh65pey4po7t122z4j@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
When working with overwritable ring buffer there's a inconvenience
problem: if perf dumps data after a long period after it starts,
non-sample events may lost, which makes following 'perf report' unable
to identify proc name and mmap layout. For example:
# perf record -m 4 -e raw_syscalls:* -g --overwrite --switch-output \
dd if=/dev/zero of=/dev/null
send SIGUSR2 after dd runs long enough. The resuling perf.data lost
correct comm and mmap events:
# perf script -i perf.data.2016061522374354
perf 24478 [004] 2581325.601789: raw_syscalls:sys_exit: NR 0 = 512
^^^^
Should be 'dd'
27b2e8 syscall_slow_exit_work+0xfe2000e3 (/lib/modules/4.6.0-rc3+/build/vmlinux)
203cc7 do_syscall_64+0xfe200117 (/lib/modules/4.6.0-rc3+/build/vmlinux)
b18d83 return_from_SYSCALL_64+0xfe200000 (/lib/modules/4.6.0-rc3+/build/vmlinux)
7f47c417edf0 [unknown] ([unknown])
^^^^^^^^^^^^
Fail to unwind
This patch provides a '--tail-synthesize' option, allows perf to collect
system status when finalizing output file. In resuling output file, the
non-sample events reflect system status when dumping data.
After this patch:
# perf record -m 4 -e raw_syscalls:* -g --overwrite --switch-output --tail-synthesize \
dd if=/dev/zero of=/dev/null
# perf script -i perf.data.2016061600544998
dd 27364 [004] 2583244.994464: raw_syscalls:sys_enter: NR 1 (1, ...
^^
Correct comm
203a18 syscall_trace_enter_phase2+0xfe2001a8 ([kernel.kallsyms])
203aa5 syscall_trace_enter+0xfe200055 ([kernel.kallsyms])
203caa do_syscall_64+0xfe2000fa ([kernel.kallsyms])
b18d83 return_from_SYSCALL_64+0xfe200000 ([kernel.kallsyms])
d8e50 __GI___libc_write+0xffff01d9639f4010 (/tmp/oxygen_root-w00229757/lib64/libc-2.18.so)
^^^^^
Correct unwind
This option doesn't aim to solve this problem completely. If a process
terminates before SIGUSR2, we still lost its COMM and MMAP events. For
example, we can't unwind correctly from the final perf.data we get from
the previous example, because when perf collects the final output file
(when we press C-c), 'dd' has been terminated so its '/proc/<pid>/mmap'
becomes empty.
However, this is a cheaper choice. To completely solve this problem we
need to continously output non-sample events. To satisify the
requirement of daemonization, we need to merge them periodically. It is
possible but requires much more code and cycles.
Automatically select --tail-synthesize when --overwrite is provided.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-16-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
If write_backward attribute is set, records are written into kernel
ring buffer from end to beginning, but read from beginning to end.
To avoid 'XX out of order events recorded' warning message (timestamps
of records is in reverse order when using write_backward), suppress the
warning message if write_backward is selected by at lease one event.
Result:
Before this patch:
# perf record -m 1 -e raw_syscalls:sys_exit/overwrite/ \
-e raw_syscalls:sys_enter \
dd if=/dev/zero of=/dev/null count=300
300+0 records in
300+0 records out
153600 bytes (154 kB) copied, 0.000601617 s, 255 MB/s
[ perf record: Woken up 5 times to write data ]
Warning:
40 out of order events recorded.
[ perf record: Captured and wrote 0.096 MB perf.data (696 samples) ]
After this patch:
# perf record -m 1 -e raw_syscalls:sys_exit/overwrite/ \
-e raw_syscalls:sys_enter \
dd if=/dev/zero of=/dev/null count=300
300+0 records in
300+0 records out
153600 bytes (154 kB) copied, 0.000644873 s, 238 MB/s
[ perf record: Woken up 5 times to write data ]
[ perf record: Captured and wrote 0.096 MB perf.data (696 samples) ]
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-15-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This patch allows following config terms and option:
Globally setting events to overwrite;
# perf record --overwrite ...
Set specific events to be overwrite or no-overwrite.
# perf record --event cycles/overwrite/ ...
# perf record --event cycles/no-overwrite/ ...
Add missing config terms and update the config term array size because
the longest string length has changed.
For overwritable events, it automatically selects attr.write_backward
since perf requires it to be backward for reading.
Test result:
# perf record --overwrite -e syscalls:*enter_nanosleep* usleep 1
[ perf record: Woken up 2 times to write data ]
[ perf record: Captured and wrote 0.011 MB perf.data (1 samples) ]
# perf evlist -v
syscalls:sys_enter_nanosleep: type: 2, size: 112, config: 0x134, { sample_period, sample_freq }: 1, sample_type: IP|TID|TIME|CPU|PERIOD|RAW, disabled: 1, inherit: 1, mmap: 1, comm: 1, enable_on_exec: 1, task: 1, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, write_backward: 1
# Tip: use 'perf evlist --trace-fields' to show fields for tracepoint events
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-14-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
There's no user of these two function outside evlist.c. Remove them from
public namespace.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: He Kuang <hekuang@huawei.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-13-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Drive the evlist->bkw_mmap_state state machine during draining and when
SIGUSR2 is received. Read the backward ring buffer in record__mmap_read_all.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-12-git-send-email-wangnan0@huawei.com
Signed-off-by: He Kuang <hekuang@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Introduce a bkw_mmap_state state machine to evlist:
.________________(forbid)_____________.
| V
NOTREADY --(0)--> RUNNING --(1)--> DATA_PENDING --(2)--> EMPTY
^ ^ | ^ |
| |__(forbid)____/ |___(forbid)___/|
| |
\_________________(3)_______________/
NOTREADY : Backward ring buffers are not ready
RUNNING : Backward ring buffers are recording
DATA_PENDING : We are required to collect data from backward ring buffers
EMPTY : We have collected data from backward ring buffers.
(0): Setup backward ring buffer
(1): Pause ring buffers for reading
(2): Read from ring buffers
(3): Resume ring buffers for recording
We can't avoid this complexity. Since we deliberately drop records from
overwritable ring buffer, there's no way for us to check remaining from
ring buffer itself (by checking head and old pointers). Therefore, we
need DATA_PENDING and EMPTY state to help us recording what we have done
to the ring buffer.
In record__mmap_read_evlist(), drive this state machine from DATA_PENDING
to EMPTY.
In perf_evlist__mmap_per_evsel(), drive this state machine from NOTREADY
to RUNNING when creating backward mmap.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: He Kuang <hekuang@huawei.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-11-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Now there's no real user of evlist->backward. Drop it. We are going to
use evlist->backward_mmap as a container for backward ring buffer.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: He Kuang <hekuang@huawei.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-10-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add backward_mmap to evlist, free it together with normal mmap.
Improve perf_evlist__pick_pc(), search backward_mmap if evlist->mmap is
not available.
This patch doesn't alloc this array. It will be allocated conditionally
in the following commits.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: He Kuang <hekuang@huawei.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-8-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
In perf_evlist__mmap_per_cpu() and perf_evlist__mmap_per_thread(), in
case of mmap failure, successfully created maps should be cleared.
Current code uses two loops calling __perf_evlist__munmap() for each
function.
This patch extracts common code to perf_evlist__munmap_nofree() and use
previous introduced decoupled API perf_mmap__munmap(). Now
__perf_evlist__munmap() can be removed because of no user.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: He Kuang <hekuang@huawei.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-7-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Insetad of saving a index into fdarray entries private field, save the
corresponding 'struct perf_mmap' pointer, and release them directly
using perf_mmap__put().
Following commits introduce multiple mmap arrays to evlist. Without this
patch, perf_evlist__munmap_filtered() is unable to retrive correct
'struct perf_mmap' pointer.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: He Kuang <hekuang@huawei.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-6-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Perf evlist will have multiple mmap arrays. Update record__mmap_read():
it should read from 'struct perf_mmap' directly.
Also, make record__mmap_read() ready to read from backward ring buffer.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: He Kuang <hekuang@huawei.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-5-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Currently, the evlist mmap related helpers and APIs accept evlist and
idx, and dereference 'struct perf_mmap' by evlist->mmap[idx]. This is
unnecessary, and force each evlist contains only one mmap array.
Following commits are going to introduce multiple mmap arrays to a
evlist. This patch refators these APIs and helpers, introduces
functions accept perf_mmap pointer directly. New helpers and APIs are
decoupled with perf_evlist, and become perf_mmap functions (so they have
perf_mmap prefix).
Old functions are reimplemented with new functions. Some of them will be
removed in following commits.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: He Kuang <hekuang@huawei.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-4-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add a 'ptr' field to fdarray->priv array.
This feature will be used by following commits, which introduce
muiltiple 'struct perf_mmap' arrays for different types of mapping.
Because of this, during fdarray__filter(), a simple 'idx' is not enough.
Add a pointer cookie that allows to directly associate a 'struct
perf_mmap' pointer to an fdarray entry.
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: He Kuang <hekuang@huawei.com>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-3-git-send-email-wangnan0@huawei.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Do it using (__CHAR_BIT__ * __SIZEOF_LONG__), simpler, works everywhere,
reduces the complexity by ditching CONFIG_64BIT, that was being
synthesized from yet another set of defines, which proved fragile,
breaking the build on linux-next for no obvious reasons.
Committer Note:
Except on:
gcc version 4.1.2 20080704 (Red Hat 4.1.2-55)
Fallback to __WORDSIZE in that case...
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160715072243.GP30154@twins.programming.kicks-ass.net
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
evsel->overwrite indicator means an event should be put into
overwritable ring buffer. In current implementation, it equals to
evsel->attr.write_backward. To reduce compliexity, remove
evsel->overwrite, use evsel->attr.write_backward instead.
In addition, in __perf_evsel__open(), if kernel doesn't support
write_backward and user explicitly set it in evsel, don't fallback
like other missing feature, since it is meaningless to fall back to
a forward ring buffer in this case: we are unable to stably read
from an forward overwritable ring buffer.
Cc: He Kuang <hekuang@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Nilay Vaish <nilayvaish@gmail.com>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: Zefan Li <lizefan@huawei.com>
Cc: pi3orama@163.com
Link: http://lkml.kernel.org/r/1468485287-33422-2-git-send-email-wangnan0@huawei.com
Signed-off-by: Wang Nan <wangnan0@huawei.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
By using 0 for base, the strtoull() detects the base automatically (see
'man strtoull').
ATM we have just one user of this function, the cpu__get_max_freq
function reading the "cpuinfo_max_freq" sysfs file. It should not get
affected by this change.
Committer note:
This change seems motivated by this discussion:
"[PATCH] [RFC V1]s390/perf: fix 'start' address of module's map"
http://lkml.kernel.org/r/20160711120155.GA29929@krava
I.e. this patches paves the way for filename__read_ull() to be used in a
S/390 related fix.
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Songshan Gong <gongss@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/1468567797-27564-4-git-send-email-jolsa@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
There are cases where further work would be needed to overcome the fact
that neither sysconf(_SC_LEVEL1_DCACHE_LINESIZE) nor
/sys/devices/system/cpu/cpu0/cache/index0/coherency_line_size are
available in some systems (Android, for instance), so bail out when such
a situation takes place.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-ho8d8g8mh0o2dri7ckcccafi@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
So far the cacheline_size is only useful for the "dcacheline" --sort
order, i.e. if that is not used, which is the norm, then the user
shouldn't care that he is running this, say, on an Android system where
sysconf(_SC_LEVEL1_DCACHE_LINESIZE) and the
/sys/devices/system/cpu/cpu0/cache/index0/coherency_line_size sysfs file
isn't available.
An upcoming patch will emit an warning only for "--sort ...,dcacheline,...".
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Chris Phlipot <cphlipot0@gmail.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-580cnkvftunyvt9n7unsholi@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
As the gcc there is producing tons of:
"warning: always_inline function might not be inlinable"
At least on android-ndk-r12/platforms/android-24/arch-arm, so, for the
time being, use this big hammer.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Chris Phlipot <cphlipot0@gmail.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-97l3eg3fnk5shmo4rsyyvj2t@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
The Bionic libc has this definition, so don't duplicate it.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Chris Phlipot <cphlipot0@gmail.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-rmd19832zkt07e4crdzyen9z@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
We need to include netinet/in.h to get the in6_addr struct definition, needed to
build it on the Android NDK:
In file included from event-parse.c:36:0:
/home/acme/android/android-ndk-r12/platforms/android-24/arch-arm/usr/include/netinet/ip6.h:82:18: error: field 'ip6_src' has incomplete type
struct in6_addr ip6_src; /* source address */
And it is the canonical way of getting IPv6 definitions, as described,
for instance, in Linux's 'man ipv6'
Doing that uncovers another problem: this source file uses PRIu64 but
doesn't include it, depending on it being included by chance via the now
replaced header (netinet/ip6.h), fix it.
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Chris Phlipot <cphlipot0@gmail.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Wang Nan <wangnan0@huawei.com>
Link: http://lkml.kernel.org/n/tip-tilr31n3yaba1whsd47qlwa3@git.kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Add a basic test case for SDT event support. This test scans an SDT
event in perftools and check whether the SDT event is correctly stored
into the buildid cache.
Here is an example:
----
$ perf test sdt -v
47: Test SDT event probing :
--- start ---
test child forked, pid 20732
Found 72 SDTs in /home/mhiramat/ksrc/linux/tools/perf/perf
Writing cache: %sdt_perf:test_target=test_target
Cache committed: 0
symbol:test_target file:(null) line:0 offset:0 return:0 lazy:(null)
test child finished with 0
---- end ----
Test SDT event probing: Ok
----
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Hemant Kumar <hemant@linux.vnet.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/146831796546.17065.1502584370844087537.stgit@devbox
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This checks whether sys/sdt.h is available or not, which is required for
DTRACE_PROBE().
We can disable this feature by passing NO_SDT=1 when building.
This flag will be used for SDT test case and further SDT events in
perftools.
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Hemant Kumar <hemant@linux.vnet.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/146831795615.17065.17513820540591053933.stgit@devbox
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Support a special SDT probe format which can omit the '%' prefix only if
the SDT group name starts with "sdt_". So, for example both of
"%sdt_libc:setjump" and "sdt_libc:setjump" are acceptable for perf probe
--add.
E.g. without this:
# perf probe -a sdt_libc:setjmp
Semantic error :There is non-digit char in line number.
...
With this:
# perf probe -a sdt_libc:setjmp
Added new event:
sdt_libc:setjmp (on %setjmp in /usr/lib64/libc-2.20.so)
You can now use it in all perf tools, such as:
perf record -e sdt_libc:setjmp -aR sleep 1
Suggested-by: Brendan Gregg <brendan.d.gregg@gmail.com>
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Hemant Kumar <hemant@linux.vnet.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/146831794674.17065.13359473252168740430.stgit@devbox
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Support @BUILDID or @FILE suffix for SDT events. This allows perf to add
probes on SDTs/pre-cached events on given FILE or the file which has
given BUILDID (also, this complements BUILDID.)
For example, both gcc and libstdc++ has same SDTs as below. If you
would like to add a probe on sdt_libstdcxx:catch on gcc, you can do as
below.
----
# perf list sdt | tail -n 6
sdt_libstdcxx:catch@/usr/bin/gcc(0cc207fc4b27) [SDT event]
sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.20(91c7a88fdf49)
sdt_libstdcxx:rethrow@/usr/bin/gcc(0cc207fc4b27) [SDT event]
sdt_libstdcxx:rethrow@/usr/lib64/libstdc++.so.6.0.20(91c7a88fdf49)
sdt_libstdcxx:throw@/usr/bin/gcc(0cc207fc4b27) [SDT event]
sdt_libstdcxx:throw@/usr/lib64/libstdc++.so.6.0.20(91c7a88fdf49)
# perf probe -a %sdt_libstdcxx:catch@0cc
Added new event:
sdt_libstdcxx:catch (on %catch in /usr/bin/gcc)
You can now use it in all perf tools, such as:
perf record -e sdt_libstdcxx:catch -aR sleep 1
----
Committer note:
Doing the full sequence of steps to get the results above:
With a clean build-id cache:
[root@jouet ~]# rm -rf ~/.debug/
[root@jouet ~]# perf list sdt
List of pre-defined events (to be used in -e):
[root@jouet ~]#
No events whatsoever, then, we can add all events in gcc to the build-id
cache, doing a --add + --dry-run:
[root@jouet ~]# perf probe --dry-run --cache -x /usr/bin/gcc --add %sdt_libstdcxx:\*
Added new events:
sdt_libstdcxx:throw (on %* in /usr/bin/gcc)
sdt_libstdcxx:rethrow (on %* in /usr/bin/gcc)
sdt_libstdcxx:catch (on %* in /usr/bin/gcc)
You can now use it in all perf tools, such as:
perf record -e sdt_libstdcxx:catch -aR sleep 1
[root@jouet ~]#
It really didn't add any events, it just cached them:
[root@jouet ~]# perf probe -l
[root@jouet ~]#
We can see that it was cached as:
[root@jouet ~]# ls -la ~/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/
total 976
drwxr-xr-x. 2 root root 4096 Jul 13 21:47 .
drwxr-xr-x. 3 root root 4096 Jul 13 21:47 ..
-rwxr-xr-x. 4 root root 985912 Jun 22 18:52 elf
-rw-r--r--. 1 root root 303 Jul 13 21:47 probes
[root@jouet ~]# file ~/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/elf
/root/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/elf: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for GNU/Linux 2.6.32, BuildID[sha1]=9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2, stripped
[root@jouet ~]# cat ~/.debug/usr/bin/gcc/9a0730e2bcc6d2a2003d21ac46807e8ee6bcb7c2/probes
%sdt_libstdcxx:throw=throw
p:sdt_libstdcxx/throw /usr/bin/gcc:0x71ffd
%sdt_libstdcxx:rethrow=rethrow
p:sdt_libstdcxx/rethrow /usr/bin/gcc:0x720b8
%sdt_libstdcxx:catch=catch
p:sdt_libstdcxx/catch /usr/bin/gcc:0x7307f
%sdt_libgcc:unwind=unwind
p:sdt_libgcc/unwind /usr/bin/gcc:0x7eec0
#sdt_libstdcxx:*=%*
[root@jouet ~]#
Ok, now we can use 'perf probe' to refer to those cached entries as:
Humm, nope, doing as above we end up with:
[root@jouet ~]# perf probe -a %sdt_libstdcxx:catch
Semantic error :* is bad for event name -it must follow C symbol-naming rule.
Error: Failed to add events.
[root@jouet ~]#
But it worked at some point, lets try not using --dry-run:
Resetting everything:
# rm -rf ~/.debug/
# perf probe -d *:*
# perf probe -l
# perf list sdt
List of pre-defined events (to be used in -e):
#
Ok, now it cached everything, even things we haven't asked it to
(sdt_libgcc:unwind):
[root@jouet ~]# perf probe -x /usr/bin/gcc --add %sdt_libstdcxx:\*
Added new events:
sdt_libstdcxx:throw (on %* in /usr/bin/gcc)
sdt_libstdcxx:rethrow (on %* in /usr/bin/gcc)
sdt_libstdcxx:catch (on %* in /usr/bin/gcc)
You can now use it in all perf tools, such as:
perf record -e sdt_libstdcxx:catch -aR sleep 1
[root@jouet ~]# perf list sdt
List of pre-defined events (to be used in -e):
sdt_libgcc:unwind [SDT event]
sdt_libstdcxx:catch [SDT event]
sdt_libstdcxx:rethrow [SDT event]
sdt_libstdcxx:throw [SDT event]
[root@jouet ~]#
And we have the events in place:
[root@jouet ~]# perf probe -l
sdt_libstdcxx:catch (on execute_cfa_program+1551@../../../libgcc/unwind-dw2.c in /usr/bin/gcc)
sdt_libstdcxx:rethrow (on d_print_subexpr+280@libsupc++/cp-demangle.c in /usr/bin/gcc)
sdt_libstdcxx:throw (on d_print_subexpr+93@libsupc++/cp-demangle.c in /usr/bin/gcc)
[root@jouet ~]#
And trying to use them at least has 'perf trace --event sdt*:*' working.
Then, if we try to add the ones in libstdc++:
[root@jouet ~]# perf probe -x /usr/lib64/libstdc++.so.6 -a %sdt_libstdcxx:\*
Error: event "catch" already exists.
Hint: Remove existing event by 'perf probe -d'
or force duplicates by 'perf probe -f'
or set 'force=yes' in BPF source.
Error: Failed to add events.
[root@jouet ~]#
Doesn't work, dups, but at least this served to, unbeknownst to the user, add
the SDT probes in /usr/lib64/libstdc++.so.6!
[root@jouet ~]# perf list sdt
List of pre-defined events (to be used in -e):
sdt_libgcc:unwind [SDT event]
sdt_libstdcxx:catch@/usr/bin/gcc(9a0730e2bcc6) [SDT event]
sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event]
sdt_libstdcxx:rethrow@/usr/bin/gcc(9a0730e2bcc6) [SDT event]
sdt_libstdcxx:rethrow@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event]
sdt_libstdcxx:throw@/usr/bin/gcc(9a0730e2bcc6) [SDT event]
sdt_libstdcxx:throw@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event]
[root@jouet ~]#
Now we should be able to get to the original cset comment, if we remove all
SDTs events in place, not from the cache, from the kernel, where it was set up as:
[root@jouet ~]# ls -la /sys/kernel/debug/tracing/events/sdt_libstdcxx/
total 0
drwxr-xr-x. 5 root root 0 Jul 13 22:00 .
drwxr-xr-x. 80 root root 0 Jul 13 21:56 ..
drwxr-xr-x. 2 root root 0 Jul 13 22:00 catch
-rw-r--r--. 1 root root 0 Jul 13 22:00 enable
-rw-r--r--. 1 root root 0 Jul 13 22:00 filter
drwxr-xr-x. 2 root root 0 Jul 13 22:00 rethrow
drwxr-xr-x. 2 root root 0 Jul 13 22:00 throw
[root@jouet ~]#
[root@jouet ~]# head -2 /sys/kernel/debug/tracing/events/sdt_libstdcxx/throw/format
name: throw
ID: 2059
[root@jouet ~]#
Now to remove it:
[root@jouet ~]# perf probe -d sdt_libstdc*:*
Removed event: sdt_libstdcxx:catch
Removed event: sdt_libstdcxx:rethrow
Removed event: sdt_libstdcxx:throw
[root@jouet ~]#
Which caused:
[root@jouet ~]# ls -la /sys/kernel/debug/tracing/events/sdt_libstdcxx/
ls: cannot access '/sys/kernel/debug/tracing/events/sdt_libstdcxx/': No such file or directory
[root@jouet ~]#
Ok, now we can do:
[root@jouet ~]# perf list sdt_libstdcxx:catch
List of pre-defined events (to be used in -e):
sdt_libstdcxx:catch@/usr/bin/gcc(9a0730e2bcc6) [SDT event]
sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event]
[root@jouet ~]#
So, these are not really 'pre-defined events', i.e. we can't use them with
'perf record --event':
[root@jouet ~]# perf record --event sdt_libstdcxx:catch*
event syntax error: 'sdt_libstdcxx:catch*'
\___ unknown tracepoint
Error: File /sys/kernel/debug/tracing/events/sdt_libstdcxx/catch* not found.
Hint: Perhaps this kernel misses some CONFIG_ setting to enable this feature?.
<SNIP>
[root@jouet ~]#
To have it really pre-defined we must use perf probe to get its definition from
the cache and set it up in the kernel, creating the tracepoint to _then_ use it
with 'perf record --event':
[root@jouet ~]# perf probe -a sdt_libstdcxx:catch
Semantic error :There is non-digit char in line number.
<SNIP>
Oops, there is another gotcha here, we need that pesky '%' character:
[root@jouet ~]# perf probe -a %sdt_libstdcxx:catch
Added new events:
sdt_libstdcxx:catch (on %catch in /usr/bin/gcc)
sdt_libstdcxx:catch_1 (on %catch in /usr/lib64/libstdc++.so.6.0.22)
You can now use it in all perf tools, such as:
perf record -e sdt_libstdcxx:catch_1 -aR sleep 1
[root@jouet ~]#
But then we added _two_ events, one with the name we expected, the other one
with a _ added, when doing the analysis we need to pay attention to who maps to
who.
And here is where we get to the point of this patch, which is to be able to
disambiguate those definitions for 'catch' in the build-id cache, but first we need
remove those events we just added:
[root@jouet ~]# perf probe -d %sdt_libstdcxx:catch
Oops, that didn't remove anything, we need to _remove_ that % char in this case:
[root@jouet ~]# perf probe -d sdt_libstdcxx:catch
Removed event: sdt_libstdcxx:catch
And we need to remove the other event added, i.e. I forgot to add a * at the end:
[root@jouet ~]# perf probe -d sdt_libstdcxx:catch*
Removed event: sdt_libstdcxx:catch_1
[root@jouet ~]#
Ok, disambiguating it using what is in this patch:
[root@jouet ~]# perf list sdt_libstdcxx:catch
List of pre-defined events (to be used in -e):
sdt_libstdcxx:catch@/usr/bin/gcc(9a0730e2bcc6) [SDT event]
sdt_libstdcxx:catch@/usr/lib64/libstdc++.so.6.0.22(ef2b7066559a) [SDT event]
[root@jouet ~]#
[root@jouet ~]# perf probe -a %sdt_libstdcxx:catch@9a07
Added new event:
sdt_libstdcxx:catch (on %catch in /usr/bin/gcc)
You can now use it in all perf tools, such as:
perf record -e sdt_libstdcxx:catch -aR sleep 1
[root@jouet ~]# perf probe -l
sdt_libstdcxx:catch (on execute_cfa_program+1551@../../../libgcc/unwind-dw2.c in /usr/bin/gcc)
[root@jouet ~]#
Yeah, it works! But we need to try and simplify this :-)
Update: Some aspects of this simplification take place in the following
patches.
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Ananth N Mavinakayanahalli <ananth@linux.vnet.ibm.com>
Cc: Brendan Gregg <brendan.d.gregg@gmail.com>
Cc: Hemant Kumar <hemant@linux.vnet.ibm.com>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/146831793746.17065.13065062753978236612.stgit@devbox
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Linus Torvalds
Dan Carpenter
Jiri Olsa
Masami Hiramatsu
Mark Rutland
Arnaldo Carvalho de Melo
Wang Nan
Peter Zijlstra
Ingo Molnar